Conventionally, attribute data is manually extracted by users out of a collected set of event data. For example, a user such as a data scientist can manually establish attributes of quantized values and determine how to organize the collected data to correspond to the attributes. However, manually extracting attribute data from collected event data is inefficient as well as laborious.